European Commission logo
INSPIRE Community Forum


Hi all,

At the INSPIRE Conference in Lissabon, we had discussions as to the meaning of the Observation ID.

The context of this discussion was the question of how the SOS is to deal with this functionality; how to avoid overflows of persisted Observations. We (UBA and AIT Austria) proposed an extension of the SOS 2.0, allowing the user to specify in the getObservation request if the Observation is to be persistent or not (or until when); thus, only those observations that need to be re-accessed are stored under their ID.

I had been of the impression that the Observation ID is generated each time a getObservation request is made to the SOS. After this point in time, it is possible to re-access the exact same Observation by getObservationByID and providing the unique ID of the observation received in the original getObservation

At the meeting in Lissabon I heard the statement that the Observation Id refers to the explicit timeseries, so more-or-less corresponds to the ObservingCapabilities, and not to the result of a previous getObservation. This was argued by the fact that the getObservationById request allows for temporal subsampling. However, taking another look at the SOS2 standard, I can’t find any parameters for getObservationById other than the ID of the observation.

Can somebody show me how to do temporal subsampling on getObservationByID, or do we need to reopen the question of “does it make sense to persist EVERY observation requested by a SOS”



Environmental Monitoring & Observations

Environmental Monitoring & Observations

Environmental Monitoring Facilities, Observations and Measurements